"Squint as we may, we fail to see the qualifications that Microsoft purports to identify in our cases."

Patent examiners give most applications only a quick look, spending on average 16 hours to 17 hours per application — nowhere near the time needed to assess whether an invention is truly new and not obvious. Worse, those hours are typically spread over two to three years, and they are interspersed with work on hundreds of other open files.

These problems could in theory be fixed with more money. But resources aren’t the only issue. The extent and quality of Patent Office review is also limited by the fact that the process is not adversarial. Indeed, the only parties involved in Patent Office review are the applicant and the applicant’s lawyers — people with an obvious incentive to see the application move forward. Contrast that with litigation, where patent plaintiffs have to square off against very motivated patent defendants.

In court, parties have to prove their case by some "standard of proof." In almost all civil cases, the standard is "preponderance of the evidence" -- meaning it is more likely than not that the facts are true. When the question is invalidating a patent, however, the U.S. Court of Appeals for the Federal Circuit decided that a defendant trying to prove a patent invalid must do so by a higher standard than normal civil cases, that of "clear and convincing" evidence. "Clear and convincing" means that the facts are "highly probable," which is a much more difficult standard to meet when trying to invalidate a patent than just a preponderance.

"As you can see from the amicus briefs, and from a letter previously sent to the Attorney General by over 260 signatories, this is an absolutely pivotal case that threatens the ability of inventors and innovative organizations to survive,"

In court, parties have to prove their case by some "standard of proof." In almost all civil cases, the standard is "preponderance of the evidence" -- meaning it is more likely than not that the facts are true. When the question is invalidating a patent, however, the U.S. Court of Appeals for the Federal Circuit decided that a defendant trying to prove a patent invalid must do so by a higher standard than normal civil cases, that of "clear and convincing" evidence. "Clear and convincing" means that the facts are "highly probable," which is a much more difficult standard to meet when trying to invalidate a patent than just a preponderance.

In court, parties have to prove their case by some "standard of proof." In almost all civil cases, the standard is "preponderance of the evidence" -- meaning it is more likely than not that the facts are true. When the question is invalidating a patent, however, the U.S. Court of Appeals for the Federal Circuit decided that a defendant trying to prove a patent invalid must do so by a higher standard than normal civil cases, that of "clear and convincing" evidence. "Clear and convincing" means that the facts are "highly probable," which is a much more difficult standard to meet when trying to invalidate a patent than just a preponderance.

Patents are essential to the modern system of innovation. Once produced, information can be transmitted at zero cost. In the absence of patent protection, would-be inventors become vulnerable to competition that would drive the value of their discovery to zero, leaving them with no compensation for the costs of producing that information in the first place.

Microsoft objects that the injunction ordered by the trial judge goes too far. (It has been put on hold until after the appeal, which is to begin Wednesday.) But injunctions are almost always ordered to prevent continuing infringement, and for good reason. To simply order money damages for future infringement would be to force i4i to license out its technology at a court-imposed price.

Just as there are good reasons not to compel citizens to sell or rent out their homes at prices set by judges, there are very good reasons in general to avoid compulsory licensing of intellectual property. Court determinations of the value of intellectual property are necessarily somewhat conjectural, yet damages awards require courts to act, in effect, as price regulators. By contrast, injunctions do not prevent a licensing deal from being done, but rather cede to the owner of the property the authority to set a price. Just as giving homeowners the right to decide whether to sell or rent out their houses does not destroy the housing market, in terrorem arguments about the death of Word under this injunction are without merit.

Meanwhile, Microsoft has vociferously argued that despite the trial judge's careful vetting of the evidence, i4i did not establish at trial a firm basis for its damages claim for past infringement. This claim about the speculative nature of past damages sits uncomfortably with Microsoft's opposition to injunctions. Given the complexity of measuring supply and demand for a unique product, it must be true that there is some empirical uncertainty about the precise level of past damages. But if patents are to have value, this uncertainty is unavoidable: A damages award is the only available remedy for infringement that has already taken place.

Protecting i4i's patent protects incentives to invent and the competitive process. In this case, the trial judge wisely offered such protection, while recognizing the court's own institutional limitations, by ordering damages for past infringement and injunctions going forward. While the decision was not a good one for Microsoft, it was clearly in the best interests of society.